Tannehill

[Dolfanalyst]

Passive aggressive comments aside, I'd be happy to, in case it wasn't obvious. (Sorry, that's two passive aggressive comments of my own .)

You state than the 28 non-failure QB's drafted in the first round have had varying levels of success. This is obviously and objectively true.

You then proceed to use this to argue that any single QB who is not seeing the greatest of success is at fault himself, since history has shown that quarterbacks "naturally" have various levels of success. This is a claim with no statistical or logical basis provided (feel free to supply it, though) and relies on an unstated (but required) argument that the varying levels of success are inherently "internal" in nature, that is, the quarterbacks fault.

One might argue that the line of reasoning is not circular in nature as much as it simply relies on an unproven hypothesis.

Let me restate your argument using

A measurement of some property of A of a population P is distributed according to some distribution D. (Success of non-failure quarterbacks chosen in the first round varies).

A(P) -> D

A measurement of A for a specific member of P, P0, returns a result that is X.

A(P0) = X

Your argument now goes: The value X is due to the inherent nature of P0.

However, at no point has it been established that the distribution D is the result of an internal property of the population P or its individual members. The only thing established is that measurements of the property A result in a distribution D when looking at P.

Thus, at best, you could argue: The value of X is due to the inherent nature of D.

This doesn't bring us anywhere, since that is just closes the circle with the original  A(P) -> D.

Take P to be a bunch of coins and A to be the number of times heads turns up after 100 flips. D is then the familiar binomial distribution.

Your claim amounts to saying that one coin that turned up 50 heads is inherently different than a coin than turned up heads 90 times. That the coin is, effectively, weighted.

In the case of NFL players, there's certainly a "weight" (skill level), but that's the very thing we're trying to establish with our candidate (Tannehill). You cannot do that simply by pointing to the distribution.

Appreciate your expounding on that in the way you did.

I'll address this part:

This is a claim with no statistical or logical basis provided (feel free to supply it, though) and relies on an unstated (but required) argument that the varying levels of success are inherently "internal" in nature, that is, the quarterbacks fault.

There is a statistical basis for that (although you're right, I hadn't yet provided it), in that when you take a group of successful quarterbacks and compare it to a group of unsuccessful quarterbacks, the between-group variation in their performance from year to year is significantly greater than the within-group variation.

If you consider year-to-year variation in their performance to represent the effect of surrounding variables (i.e., otherwise their internal level of ability would make them play the same way every year, and there would be no such year-to-year variation), the finding suggests that internal factors, rather than external ones, drive the bus in this area.

To put it more simply, the better quarterbacks hover around a significantly higher level of performance than the worse ones.  What that suggests is that a better quarterback would've performed significantly better than Tannehill amidst the same surroundings on the 2015 Miami Dolphins.  Conversely, a worse quarterback would've performed worse under those same circumstances.

The take-home message is that quarterbacks aren't hapless victims of their surroundings.  They have an internal level of ability that can compensate for and overcome their surroundings (or not).

[Rich]

What that suggests TO ME AND ONLY ME is that a better quarterback would've performed significantly better than Tannehill amidst the same surroundings on the 2015 Miami Dolphins.  Conversely, a worse quarterback would've performed worse under those same circumstances.

Fixed it for ya.

[Brian Fein]

There is still the possibility that he would play much better behind a better offensive line.  There is just no quantitative evidence to support that assertion.  The support for that assertion rests entirely in the land of theory and subjective opinion.

You don't need to have quantitative evidence to predict the future on a football field.  It can only be verified by experimental methodology.  So, let's go get a stellar o-line and prove the hypothesis.

[fyo]

You don't need to have quantitative evidence to predict the future on a football field.  It can only be verified by experimental methodology.  So, let's go get a stellar o-line and prove the hypothesis.

You heard it Tannenbaum! Now go do it! We need the data.

[Rich]

You heard it Tannenbaum! Now go do it! We need the data.

Really, all it would take is finding two serviceable guards and some decent depth. We have the starting tackles and center in place.

[Dolfanalyst]

Really, all it would take is finding two serviceable guards and some decent depth. We have the starting tackles and center in place.

I'm not trying to be argumentative in saying this, but based on the correlational evidence that exists between offensive lines and quarterback play in the league, you'd be better off betting money that the presence of those changes in the offensive line wouldn't improve Tannehill's performance significantly.

It's very easy to theorize on a message board that those changes would improve Tannehill's performance, but anyone who had a gun held to his head and was told to bet all the money he had on the issue or else would be wise to do the quantitative research and base his bet on that, rather than theory and observation.  You're much more likely to win that way.

[Rich]

I'm not trying to be argumentative in saying this, but based on the correlational evidence that exists between offensive lines and quarterback play in the league, you'd be better off betting money that the presence of those changes in the offensive line wouldn't improve Tannehill's performance significantly.

It's very easy to theorize on a message board that those changes would improve Tannehill's performance, but anyone who had a gun held to his head and was told to bet all the money he had on the issue or else would be wise to do the quantitative research and base his bet on that, rather than theory and observation.  You're much more likely to win that way.

So your solution to the offensive line woes would be to not change the guards.

Good, we'll mark you down as being on that side of the fence.

[Dolfanalyst]

So your solution to the offensive line woes would be to not change the guards.

I don't believe there are "offensive line woes" that are significantly greater than the average team's, so yes, I would take the limited amount of resources available for personnel improvement and place them elsewhere.

Good, we'll mark you down as being on that side of the fence.

I'm not really interested in the fence, or who's on either side of it.  That sounds like "message board politics" to me, and I don't practice those.

[Rich]

I don't believe there are "offensive line woes"

Maybe you should watch the games and stop staring at the spreadsheet.

[Dave Gray]

Donafanalyst, do you feel that the "eye test" has any value?

I respect stats, but when I watch the game, I see a lack of protection.

I believe that our coaches also see this and call a game with this in mind.  This skews the stats, since the coordinators are almost forfeiting sections of a playbook.

[Dolfanalyst]

Maybe you should watch the games and stop staring at the spreadsheet.

Watching the Dolphins' games tells us nothing about how the Dolphins' offensive line compares to those of the other teams in the league.

The only people I'm aware of who watch every team's games and rate their offensive lines based on those observations (the PFF folks) have told us the Dolphins' pass blocking isn't significantly worse than that of the average team.

[Rich]

Watching the Dolphins' games tells us nothing about how the Dolphins' offensive line compares to those of the other teams in the league.

The only people I'm aware of who watch every team's games and rate their offensive lines based on those observations (the PFF folks) have told us the Dolphins' pass blocking isn't significantly worse than that of the average team.

Really?

Here is PFF's summary of the Dolphins offensive line, which they rated as 30th in the league.

Stud: You almost feel bad for Mike Pouncey (80.0) at times, who looks so much better than his line teammates.

Dud: Pick a guard, any guard. Well, any guard named Dallas Thomas (36.3) or Jamil Douglas (28.7). Neither man has come out of the first eight weeks with any real credit.

Summary: The offensive line looked like a weakness heading into the season, and so it has proved. Injuries to Ja’Wuan James and , haven’t helped, but even those two being fully healthy couldn’t overcome the play at the guard spots.

Hmmm... Any guard is a dud, it looks like a weakness...

Those don't sound like ringing endorsements.

[Rich]

Donafanalyst, do you feel that the "eye test" has any value?

I respect stats, but when I watch the game, I see a lack of protection.

I believe that our coaches also see this and call a game with this in mind.  This skews the stats, since the coordinators are almost forfeiting sections of a playbook.

To summarize his response before he gives it, your eyes are lying to you.

[Dolfanalyst]

Donafanalyst, do you feel that the "eye test" has any value?

Very little, when we're watching the team we care deeply about with much greater interest and intent than we are the rest of the league's games, and doing no systematic comparisons among those teams based on our observations.  In fact, I think the "eye test" can have negative value, in that it invites confirmation bias under those circumstances.

I respect stats, but when I watch the game, I see a lack of protection.

And I don't disrespect anyone's personal opinion based on their observations, but if you believe the team suffers from poor protection, you're very likely to "see" that when you watch the team's games, i.e., confirmation bias.  Here's more on that concept:

Confirmation bias, also called confirmatory bias or myside bias, is the tendency to search for, interpret, favor, and recall information in a way that confirms one's beliefs or hypotheses, while giving disproportionately less consideration to alternative possibilities. It is a type of cognitive bias and a systematic error of inductive reasoning. People display this bias when they gather or remember information selectively, or when they interpret it in a biased way. The effect is stronger for emotionally charged issues [like Ryan Tannehill's success] and for deeply entrenched beliefs. People also tend to interpret ambiguous evidence as supporting their existing position.

https://en.wikipedia.org/wiki/Confirmation_bias

That's simply human nature.  We're all prone to it, myself included of course.

I believe that our coaches also see this and call a game with this in mind.  This skews the stats, since the coordinators are almost forfeiting sections of a playbook.

And that's a viable hypothesis.

An equally viable hypothesis is that the coordinators are tailoring their gameplans around Ryan Tannehill's weaknesses, one of which is escaping pressure with movement.

How do we know which hypothesis is correct?

If we believe the protection is poor, and we're experiencing confirmation bias when we watch the games, then we're likely to "confirm" the first hypothesis in our own minds, and "disconfirm" the second one, without even being aware of it.

Again, we have to be very careful when we very badly want to see a certain thing (i.e., Ryan Tannehill's success).  Under those conditions we're far more prone to make conclusions based on distorted perceptions.  We're only human, after all.

Someone here made the comment earlier that the prevailing belief once upon a time was that Chad Henne was only "a number-one receiver" away from greatness was similar to how some folks are viewing Ryan Tannehill.

That prevailing belief was wrong, which should humble us about the reliability of similar perceptions we may have.  This one about Ryan Tannehill and his offensive line is very, very similar.

[Brian Fein]

The only people I'm aware of who watch every team's games and rate their offensive lines based on those observations (the PFF folks) have told us the Dolphins' pass blocking isn't significantly worse than that of the average team.

Here's where you go wrong - assuming that PFF's rankings are normally distributed and that your assertion of "not more than one standard deviation away" has any significance.  That just speaks to a large standard deviation in a sample set that is too small to be statistically significant.  

You can not apply statistical analysis to PFF's ranking and rating and draw any reasonable conclusion.  To do so would essentially say that every offensive line in the league is equivalent in terms of performance.  You're a smart enough guy to know that this statement is false, right?